European Journal of Internal Medicine 24 (2013) 536–540

Contents lists available at ScienceDirect

European Journal of Internal Medicine

j ourna l homepage: www.e lsev ie r .com/ locate /e j im

Original article

Trends in the burden of infectious disease hospitalizations among theelderly in the last decade

Walid Saliba a,b,⁎, Anna Fediai c, Hana Edelstein d, Arie Markel b,c, Raul Raz b,d

a Department of Internal Medicine C, Ha'emek Medical Center, Afula, Israelb Bruce Rappaport Faculty of Medicine, Technion—Israel Institute of Technology, Haifa, Israelc Department of Internal Medicine A, Ha'emek Medical Center, Afula, Israeld Infectious Disease Unit, Ha'emek Medical Center, Afula, Israel

⁎ Corresponding author at: Department of Internal MediAfula 18101, Israel. Tel.: +972 4 6495132; fax: +972 4 64

E-mail address: saliba_wa@clalit.org.il (W. Saliba).

0953-6205/$ – see front matter © 2013 European Federhttp://dx.doi.org/10.1016/j.ejim.2013.06.002

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 9 February 2013Received in revised form 25 May 2013Accepted 8 June 2013Available online 28 June 2013

Keywords:Infectious diseasesElderlyHospitalizationTrend

Background: Infectious disease is a leading cause of hospitalization. We investigated trends in infectious dis-ease hospitalizations among the elderly in the last decade.Methods: A total of 81,077 hospitalizations of elderly patients between 2001 and 2010 were available on thecomputerized database of the Ha'emek Medical Center, Israel. The proportion of hospitalizations attributableto infectious diseases was calculated.Results: Overall, lower respiratory tract infection (LRTI) accounted for 41.0% of hospitalizations attributable to in-fectious diseases followed by kidney, urinary tract and bladder infections (UTI) (21.4%), upper respiratory tractinfections (URTI) (10.2%), and hepatobiliary tract infections (9.8%).The proportion of hospitalizations attributable to infectious diseases increased by 14.2% during the studyperiod, rising from 16.9% in 2001 (1023 infectious disease hospitalizations of a total of 6043 hospitaliza-

tions) to 19.3% in 2010 (1907 infectious disease hospitalizations of a total of 9876 hospitalizations) (P fortrend b 0.001). A significant increasing trend persisted after adjustment for age, ethnicity, and season,resulting in an increase from 16.9% in 2001 to 18.8% in 2010 (P for trend = 0.001). A significant increasingtrend was observed in males (P for trend b 0.001) and a borderline significant trend was observed in fe-males (P for trend = 0.062). The proportion of hospitalizations attributable to infectious diseases washigher in males and increased with age. LRTI and URTI were the major contributors to the increasingtrend (P for trend = 0.018 and b0.001, respectively).Conclusions: This study shows an increasing trend in infectious disease hospitalizations among theelderly in the last decade. Public health measures are needed to reduce infectious diseasehospitalizations.

© 2013 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.

1. Introduction

Infectious disease is associated with substantial morbidity andmortality and remains a leading cause of visits to ambulatory clinicsand hospitalization [1–4]. Infectious diseases accounted for 19% ofvisits to physicians in the United States [3]. In 1994, infectiousdiseases accounted for 23.3% of hospitalizations among AmericanIndians and older adult Alaskan natives [5], 11% of hospitalizationsamong American whites, and 12% of hospitalizations among blacksand all other races [1].

Previous studies indicate an increasing trend in the burden of in-fectious disease hospitalizations, as reflected by an increase in theproportion of hospitalizations attributable to infectious diseases[1,5,6]. In the last few years, the overall rate of hospitalizations in

cine C, Ha'emekMedical Center,95134.

ation of Internal Medicine. Publishe

the United States declined by approximately 33%, while the rate ofhospitalizations for infectious diseases declined less steeply inthose aged b65 and increased in those aged ≥65 years [1,7]. Conse-quently, the proportion of hospitalizations attributable to infectiousdiseases increased [1]. This may be explained by the improvement inthe treatment of chronic diseases resulting in less need for hospital-ization for these conditions. In addition, the aging of the population[7,8] and the increasing prevalence of elderly patients with comorbidconditions has resulted in a population that is more susceptible to in-fectious diseases [1,6,7]. Moreover, because of comorbidities, elderlysubjects are more likely to visit the emergency department, a findingassociated with a three-fold increased risk of acute infection [9]. An-other important contributor is the widespread use of antibiotics andthe emergence of highly resistant pathogens [10].

Recent data concerning infectious disease hospitalizations is lack-ing in Israel. This study describes the trend in the burden of hospital-izations attributable to infectious diseases among the elderly duringthe last decade in Israel.

d by Elsevier B.V. All rights reserved.

Table 1Characteristics of elderly subjects hospitalized at the Ha'emek Medical Center duringthe years 2001–2010 stratified by the causes of hospitalization (infectious versusnoninfectious).

Variable All(n = 81,077)

Type of admission Pvalue

Infectious(n = 15,195)

Noninfectious(n = 65,882)

Gender b0.001Males 40,569 (50%) 7858 (51.7%) 32,709 (49.6%)Females 40,508 (50%) 7335 (48.3%) 33,173 (50.4%)

Ethnicity b0.001Jews 65,597 (80.9%) 11,914 (78.4%) 53,683 (81.5%)Arabs 15,480 (19.1%) 3281 (21.6%) 12,199 (18.5%)

Age group b0.00165–74.9 years 36,506 (45.0%) 5756 (37.9%) 30,750 (46.7%)75–84.9 years 32,699 (40.3%) 6329 (41.7%) 26,370 (40.0%)≥85 years 11,872 (14.7%) 3110 (20.5%) 8762 (13.3%)

Season 0.004Winter 21,891 (27.0%) 4192 (27.6%) 17,699 (26.9%)Spring 20,250 (25.0%) 3859 (25.4%) 16,391 (24.9%)Summer 19,666 (24.3%) 3516 (23.1%) 16,150 (24.5%)Autumn 19,270 (23.7%) 3628 (23.9%) 15,642 (23.7%)

Ward of admission b0.001Internal medicine 54,471 (67.3%) 11,886 (78.2%) 42,585 (64.4%)General surgery 10,740 (13.2%) 1949 (12.8%) 8791 (13.3%)Urology 4964 (6.1%) 958 (6.3%) 4006 (6.1%)Orthopedic 5071 (6.3%) 271 (1.8%) 4800 (7.3%)Gynecology 1009 (1.2%) 43 (0.3%) 966 (1.5%)Intensive care 4822 (5.9%) 88 (0.6%) 4734 (7.2%)

Duration of admission(days)Mean ± SD 4.6 ± 5.8 5.4 ± 6.5 4.4 ± 5.6 b0.001

Charlson comorbidityindexMedian(inter-quartile range)

1 (0–2) 1 (0–2) 1 (0–2) b0.001

537W. Saliba et al. / European Journal of Internal Medicine 24 (2013) 536–540

2. Materials and methods

2.1. Study population and data source

The study was conducted at the Ha'emek Medical Center, locatedin the northeastern region of Israel, which serves a population ofmore than 500,000 persons. Ha'emek Medical Center is one of a num-ber of hospitals in the region and does not have specific referralsub-specialties. Referral to any of the hospitals in the region is depen-dent on the preference of the primary care physician, the patient, andthe patient's family. This study was approved by the local institutionalreview board and ethics committee.

We included all admissions of elderly patients (≥65 years old at thetime of hospitalization) at the Ha'emek Medical Center between 2001and 2010 to any of the following wards: internal medicine, intensivecare units, general surgery, urology, orthopedics, and gynecology.

Hospitalization data are available from a computerized database andincludes reason for hospitalization, date of hospitalization, date of dis-charge or death, age at the time of hospitalization, gender, ethnicgroup, and a calculated Charlson comorbidity index for each admission.

If the patient was transferred to another ward during the hospital-ization period, only the hospitalization in the first ward was consid-ered in this study.

2.2. Infectious disease classification

The causes or reasons for hospitalization were classified into twocategories: infectious causes and noninfectious causes. Infectious causeswere grouped into the following categories according to the site or thetype of infection using the ICD-9 coding: lower respiratory tract infec-tion, upper respiratory tract infection, kidney, urinary tract and bladderinfection, enteric infection, viral central nervous system infection, men-ingitis, cellulitis, tuberculosis, septicemia, hepatobiliary disease, infec-tions of the heart, abdominal and rectal infection, infection due tointernal prosthetic device, implant and graft, postoperative infection,osteomyelitis, periostitis, and other infections involving bones, and in-flammatory disease of female pelvic organs.

2.3. Definition of terms

The proportion of hospitalizations attributable to infectious dis-eases was calculated by dividing the number of hospitalizations dueto infectious diseases in each year by the total numbers of hospitaliza-tions in the same year.

The Charlson comorbidity index is a weighted index that takes intoaccount the number and the seriousness of comorbid conditions [11].To study trends in the severity of comorbid conditions, we used the tra-ditional Charlson comorbidity index grouped into four categories: score0, no comorbid conditions; score 1–2, mild; score 3–4, moderate; score≥5, severe [6,11].

2.4. Statistical analyses

Continuous variables were presented as means and standard devia-tions (SD) ormedianswith the inter-quartile range (IQR) as appropriate.Categorical data were presented as proportions. The comparison of con-tinuous variables between two categorieswas performedwith Student'st-test or the Mann–Whitney test as appropriate. The association be-tween categorical variables was tested with the Chi-square test.

The proportion of hospitalizations attributable to infectiousdiseases was adjusted for age, ethnicity, and season by means of thedirect adjustment method, using the population of the year 2001 asthe standard reference population.

Trends in the adjusted proportions during the study period(2001–2010) were studied with logistic regression. The adjusted Pfor trend was estimated by including the year of the study period as

a continuous variable in the model. The method of generalized esti-mating equations (GEE) was used to account for correlations of hospi-talizations from the same subject.

A P-value of less than 0.05 for the two-tailed test was consideredstatistically significant. All statistical analyses were performed usingSPSS 18.0 (SPSS Inc., Chicago).

3. Results

A total of 81,284 hospitalizations of elderly patients were identi-fied during the study period in the analyzed wards of the Ha'emekMedical Center. Because the cause of hospitalization was missing in207 hospitalizations, 81,077 hospitalizations were included in thefinal analyses. Overall, 15,195 (18.7%) of the hospitalizations were at-tributable to infectious diseases.

Compared with subjects hospitalized with noninfectious diseases,the subjects with infectious diseases were more likely to be males, pa-tients of Arab origin, of older age, and to have a longer duration ofhospitalization (P value b 0.001 for each) (Table 1).

3.1. Infectious disease hospitalizations

Overall, lower respiratory tract infections (LRTI) accounted for41.0% of hospitalizations attributable to infectious diseases, followedby kidney, urinary tract and bladder infections (UTI) (21.4%), upperrespiratory tract infections (URTI) (10.2%), and hepatobiliary tract in-fections (9.8%). LRTI accounted for 49.4%, 45.9%, 32.1%, and 34.8% ofinfectious disease hospitalizations in the winter, spring, summer,and autumn, respectively (P value b 0.001).

During the study period, there was a significant increasing trend inthe crude proportion of hospitalizations attributable to infectiousdiseases: from 16.9% (1023 infectious disease hospitalizations of a totalof 6043 hospitalizations) in 2001 to 19.3% (1907 infectious disease

A

10%

12%

14%

16%

18%

20%

22%

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Year

Pro

po

rtio

n o

f h

osp

ital

izat

ion

s d

ue

to in

fect

iou

s d

isea

ses

Males Females

B

10%

12%

14%

16%

18%

20%

22%

24%

26%

28%

30%

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Year

Pro

po

rtio

n o

f h

osp

ital

izti

on

du

e to

Infe

ctio

us

dis

ease

s

65-74.9 75-84.9 >=85

Fig. 2. A. Trends in the age-, ethnicity-, and season-adjusted proportions of hospitaliza-tions attributable to infectious diseases stratified by gender. B. Trends in the age-specificproportions of hospitalizations attributable to infectious diseases; Ha'emek Medical Cen-ter (2001–2010).

538 W. Saliba et al. / European Journal of Internal Medicine 24 (2013) 536–540

hospitalizations of a total of 9876 hospitalizations) in 2010, reflecting anincrease of 14.2% (P for trend b 0.001). The age-adjusted proportion ofhospitalizations attributable to infectious diseases was very similar:16.9% in 2001 and 19.0% in 2010 (P for trend b 0.001) (Fig. 1). The in-creasing trend persisted after further adjustments for ethnicity and sea-sonality: 16.9% in 2001 and 18.8% in 2010 (P for trend = 0.001) (Fig. 1).A significant increasing trend in age-, ethnicity-, and season-adjustedproportions of infectious disease hospitalizations was observed inmales (P for trend = 0.001), while a borderline significant increasingtrend was observed in female patients (P for trend = 0.062) (Fig. 2A).An increasing trend in the proportions of infectious disease hospitaliza-tions was observed in subjects aged 65–74.9 years (P for trend b 0.001)and in subjects aged 75–84.9 years (P for trend = 0.024). Subjects aged≥85 years did not show a specific pattern. In this group of patients, thelowest proportion was observed in 2001 (22.9%) and the highest pro-portion (28.6%) in 2004; in 2010, the proportion of infectious diseasehospitalizations was (24.5%) (P for trend = 0.623) (Fig. 2B).

The age-, ethnicity-, and season-adjusted proportions of hospitaliza-tions attributable to infectious diseases were consistently higher inmales than in females in most of the study years (Fig. 2A). Overall,7886 (19.4%) of hospitalizations in males and 7335 (18.1%) of hospital-izations in females were attributable to infectious diseases (P b 0.001).The proportion of hospitalizations attributable to infectious diseases in-creased significantly with increasing age (Fig. 2B). Overall, 5756 (15.8%)hospitalizations in patients aged 65–74.9 years, 6329 (19.4%) hospitali-zations in patients aged 75–84.9 years, and 3110 (26.2%) hospitaliza-tions in patients aged ≥85 years were attributable to infectiousdiseases (P b 0.001). In a multivariate analysis, age, gender, ethnicityand seasonality were independently associated with infectious diseasehospitalizations (Table 2).

An increasingly significant trendwas observed in hospitalizations ininternal medicine wards, with infectious diseases being responsible for18.3% of hospitalizations in 2001 and 22.0% in 2010, reflecting an in-crease of 20.2% (P for trend b 0.001). A decreasing trend in the propor-tion of hospitalizations attributable to infectious diseases was observedin hospitalizations in the orthopedic ward (P for trend = 0.003). Theproportion of hospitalizations attributable to infectious diseases in theother departmentswere either stable or did not show a specific pattern.

The age-, ethnicity-, and season-adjusted proportions of hospi-talizations attributable to LRTI and URTI showed a significant trend(P for trend = 0.018 and b0.001, respectively). The proportion ofhospitalizations attributable to UTI and hepatobiliary tract infec-tions showed an almost stable pattern during the study period(Fig. 3).

15%

16%

17%

18%

19%

20%

21%

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Year

Pro

po

rtio

n o

f h

osp

ital

izat

ion

s d

ue

to in

fect

iou

s d

isea

ses

Crude Age-adjusted Age, ethnicity, and season adjusted

Fig. 1. Trends in crude, age-adjusted, and age-, ethnicity-, and season-adjusted propor-tions of hospitalizations attributable to infectious diseases; Ha'emek Medical Center(2001–2010).

3.2. Comorbid conditions

The average score of the Charlson comorbidity index was similar inpatients hospitalized with infectious diseases and with noninfectiousdiseases (Table 1). The severity of comorbid conditions among patientshospitalized because of infectious diseases was higher in 2010 than in2001 in all age groups (Fig. 4). The Charlson comorbidity index was in-dependently associated with infectious disease hospitalization in themultivariate analysis (Table 2).

4. Discussion

We observed a significant increasing trend in the proportion ofhospitalizations attributable to infectious diseases in the last decade.

The proportion of hospitalizations attributable to infectious diseasesincreased by 14.2% from 2001 to 2010. As shown in Fig. 1, the increasingtrend in the proportion of hospitalizations attributable to infectious dis-eases persisted after adjustment for age, ethnicity, and season. The ad-justed proportions of infectious disease hospitalizations are similar tothe crude proportions that more accurately reflect the true burden ofinfectious disease hospitalizations. Although the observed increasedprevalence of comorbid conditions (Fig. 4) may partially explain ourfindings, it may also merely represent an increased awareness on thepart of the physician for recording comorbid conditions. In line withour findings, previous studies from the USA revealed an increase in theproportion of hospitalizations attributable to infectious diseases [1,5].However, it should be emphasized that these studies were based on adifferent healthcare system with different population demographics.The published estimates are therefore not directly comparable.

The proportion of hospitalizations attributable to infectious dis-eases is higher in the older age groups and in males. Similar age and

Table 2Multivariate analysis; adjusted odds ratio (OR and 95% confidence interval) for infec-tious disease hospitalizations; Ha'emek Medical Center (2001–2010).

Variable OR (95% CI) P value

GenderMales 1.168 (1.126–1.211) b0.001Females Reference

EthnicityJews ReferenceArabs 1.220 (1.167–1.276) b0.001

Age group65–74.9 years Reference75–84.9 years 1.250 (1.200–1.301) b0.001≥85 years 1.794 (1.706–1.888) b0.001

SeasonWinter 1.072 (1.019–1.127) 0.007Spring 1.070 (1.017–1.127) 0.009Summer ReferenceAutumn 1.074 (1.019–1.131) 0.008

Ward of admissionInternal medicine 1.962 (1.878–2.049) b0.001All other wards Reference

Charlson comorbidity indexNo comorbidities ReferenceMild (1–2 comorbidities) 1.103 (1.058–1.150) b0.001Moderate (3–4 comorbidities) 0.983 (0.926–1.043) 0.566Severe (≥5 comorbidities) 0.677 (0.616–0.743) b0.001Duration of admission (for 1 day increase) 1.031 (1.028–1.034) b0.001Year of the study (for 1 year increase) 1.010 (1.004–1.017) 0.002

539W. Saliba et al. / European Journal of Internal Medicine 24 (2013) 536–540

gender differences in infectious disease hospitalization have beenreported previously [1,6,7,12]. This may be partially explained bythe higher prevalence of comorbid conditions in these groups,resulting in an increased susceptibility to infectious diseases [13].

LRTI is the most frequently diagnosed infectious disease andaccounted for 41.0% of infectious disease hospitalizations. This findingis consistent with the findings of other studies; LRTI accounted for37.0% of all infectious disease hospitalizations in the USA, and 46.1% ofinfectious diseases among older adults [1,7]. The increasing trend inthe infectious disease hospitalization rate is mainly due to LRTI andURTI hospitalizations. Interestingly, this occurred in spite of better vac-cine coverage for vaccine-preventable respiratory infections. Data fromIsrael show an increasing trend in the pneumococcal vaccination rate inthe elderly [14]. While official data on influenza vaccination in the el-derly are lacking, much has been undertaken to increase the influenzavaccination rate in the last decade. These findings may therefore beexplained by the increasing prevalence of elderly patients with comor-bid conditions resulting in an increased susceptibility to respiratory

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Year

Pro

po

rtio

n o

f h

osp

ital

izat

ion

s d

ue

to in

fect

iou

s d

isea

se

URTI LRTI UTI Hepatobiliary

Fig. 3. Trends in age-, ethnicity-, and season-adjusted proportions of hospitalizations at-tributable to specific infectious disease groups; Ha'emek Medical Center (2001–2010).

infections or by a propensity to treat these conditions in a hospital set-ting. An increasing trend in LRTI has been reported previously [6], whileanother study found a decreasing trend in URTI hospitalizations [1].

This study shows that the significant increasing trend is confined tointernalmedicinewards. Thismay be explained by the nature of the twoconditions that are responsible for this trend, namely LRTI and URTI,both of which are grounds for admission to internal medicine wards.

This study has several limitations. We cannot estimate the infec-tious disease hospitalization rate representing the proportion of thepopulation requiring hospitalization for infectious diseases indepen-dently from the rate of noninfectious disease hospitalization. This isnot possible because the number of infectious disease admissions atHa'emek Medical Center constitutes only a part of all of the infectiousdisease admissions in the area. Our hospital is not the only hospital inthe region and referral to any of the hospitals is dependent on thepreference of the patient and the primary care physician.

In addition, using the reason for hospitalization does not take intoaccount healthcare-associated infections (HAI) that may have oc-curred during hospitalization, which are associated with increasedof length of stay in the hospital and increased mortality. Much prog-ress has been made over the past decade in HAI prevention and con-trol. While these efforts may have resulted in a partial overestimationof the detected trend of infectious disease hospitalization, it is lesslikely to be responsible for the entire trend.

Another limitation of this study is the lack of data on the differentia-tion between viral and bacterial etiology of respiratory tract infections.A recent study from Ha'emek Medical Center investigated the etiologyof community-acquired lower respiratory tract infections in hospitalizedpatients [15]. All participants underwent an extensive workup, includingblood and sputumcultures, serology, and respiratorymultiplex-PCR fromnasopharyngeal swabs. At least one pathogen was found in 66.7% of thecases, and no pathogens were found in 33.7% of the cases. A viral patho-genwas identified in 33.3% of the cases, typical bacteriawere identified in18.3% of the cases, and atypical bacterial were identified in 52.4% of thecases [15].

The data from this study come from a single medical center, whichmay affect the external validity of the results; however, the results arein line with those of previous studies. In addition, the observed trendmay result from changes that may have occurred during the study pe-riod, including hospitalization criteria, new diagnostic methods, andadmission capacity of the hospital.

The burden of hospitalizations attributable to infectious diseaseshas increased among the elderly in the last decade. The proportionof hospitalizations attributable to infectious diseases is likely to in-crease further in the future as a result of the expected aging of thepopulation and the increasing prevalence of comorbid conditionsamong the elderly. As observed in this study, the burden of infectiousdisease hospitalizations is further exaggerated by the longer durationof hospitalizations. Hence, efforts should be made to implementstricter guidelines for hospitalization and the referral of the elderlyto the emergency department together with developing novel anti-microbial agents and a wiser use of antibiotics concordant withguidelines. Moreover, because the increasing trend in the proportionof hospitalizations attributable to infectious diseases resulted mainlyfrom the trend in LRTI and URTI, we strongly recommend more ex-tensive public health interventions to increase the influenza andpneumococcal vaccination rate.

Learning points

• The burden of infectious disease hospitalizations has increasedamong the elderly in the last decade

• Lower and upper respiratory tract infections are the major contribu-tors to the increasing trend of infectious diseases among the elderly

• Lower and upper respiratory tract infections account for more thanhalf of the hospitalizations attributable to infectious diseases

0%

10%

20%

30%

40%

50%

60%

70%

No

Mild

Mod

erat

e

Sev

ere

No

Mild

Mod

erat

e

Sev

ere

No

Mild

Mod

erat

e

Sev

ere

65-74.9 75-84.9 ≥85

Charlson comorbidity index severity by age group

Per

cen

tag

e

2001 2010

Fig. 4. Comparison of Charlson comorbidity index severity between 2001 and 2010 in patients hospitalized because of infectious diseases stratified by age groups; Ha'emek MedicalCenter (2001–2010).

540 W. Saliba et al. / European Journal of Internal Medicine 24 (2013) 536–540

Conflict of interests

No conflict of interests or financial disclosures were reported bythe authors of this paper.

References

[1] Simonsen L, ConnLA, PinnerRW, TeutschS. Trends in infectiousdiseasehospitalizationsin the United States, 1980–1994. Arch Intern Med 1998;158:1923–8.

[2] Armstrong GL, Conn LA, Pinner RW. Trends in infectious disease mortality in theUnited States during the 20th century. JAMA 1999;281:61–6.

[3] Armstrong GL, Pinner RW. Outpatient visits for infectious diseases in the UnitedStates, 1980 through 1996. Arch Intern Med 1999;159:2531–6.

[4] SonnenblickM, RavehD, Gratch L, Yinnon A. Clinical and demographic characteristicsof elderly patients hospitalised in an internal medicine department in Israel. Int J ClinPract 2007;61:247–54.

[5] Holman RC, Curns AT, Singleton RJ, Sejvar JJ, Butler JC, Paisano EL, et al. Infectiousdisease hospitalizations among older American Indian and Alaska Native adults.Public Health Rep 2006;121:674–83.

[6] Trotter CL, Stuart JM,George R,Miller E. Increasinghospital admissions for pneumonia,England. Emerg Infect Dis 2008;14:727–33.

[7] Curns AT, Holman RC, Sejvar JJ, Owings MF, Schonberger LB. Infectious diseasehospitalizations among older adults in the United States from 1990 through2002. Arch Intern Med 2005;165:2514–20.

[8] Day JC. Population projection of the United States by age, sex, and Hispanic origin:1995 to 2050. Washington DC: US Government Printing Office; 1996 .

[9] Quach C, McArthur M, McGeer A, Li L, Simor A, Dionne M, et al. Risk of infectionfollowing a visit to the emergency department: a cohort study. CMAJ 2012;184:E232–9.

[10] Nimri LF, BatchounR. Community-acquired bacteraemia in a rural area: predominantbacterial species and antibiotic resistance. J Med Microbiol 2004;53:1045–9.

[11] CharlsonME, Pompei P, Ales KL,MacKenzie CR. A newmethod of classifying prognosticcomorbidity in longitudinal studies: development and validation. J Chronic Dis1987;40:373–83.

[12] Christensen KL, Holman RC, Steiner CA, Sejvar JJ, Stoll BJ, Schonberger LB. Infectiousdisease hospitalizations in the United States. Clin Infect Dis 2009;49:1025–35.

[13] Kappert K, Böhm M, Schmieder R, Schumacher H, Teo K, Yusuf S, et al. Impact ofsex on cardiovascular outcome in patients at high cardiovascular risk: analysisof the Telmisartan Randomized Assessment Study in ACE-Intolerant Subjectswith Cardiovascular Disease (TRANSCEND) and the Ongoing Telmisartan Aloneand in Combination with Ramipril Global End Point Trial (ONTARGET). Circula-tion 2012;126:934–41.

[14] CohenAD,Dreiher J, GoldfrachtM,YakovsonO, LiebermanN,Weiss D, et al. The effectof the introduction of pneumococcal vaccination for adults as a quality measure onvaccination rates in Clalit Health Services. Harefuah 2011;150:578–82.

[15] Shibli F, Chazan B, Nitzan O, Flatau E, Edelstein H, Blondheim O, et al. Etiology ofcommunity-acquired pneumonia in hospitalized patients in northern Israel. IsrMed Assoc J 2010;12:477–82.